A REAL LIFE WEB BASED MARKETING OPTIMIZATION FRAMEWORK with EXTERNAL DATA

Abstract

Big data and data science studies in recent years are booming exponentially, parallel to the data collected and increased processing speeds. As an inevitable consequence, most of the web-based companies are migrating their business models to novel technologies based on the big data and data science research. This paper is based on a real life experience based on one of the web stores with highest volume sales in Turkey. The project was building a data science model on big data technologies to make estimations based on the external data, such as weather conditions, customer demography, news at newspapers, current product alternatives, financial facts (like currency exchange rate or stock market values) and most importantly the sentimental analysis and opinion mining on social network, blogs and news. In the paper, details of problems and possible solution alternatives and methodology for problem solving and solutions and outcomes of the study are explained in the given order.